This part should be significantly shorter than the last one. In this, I will review my sources for materials. The reviews will be pretty short, as purchasing items are typically pretty straightforward...even in cases of small amounts of uranium...*cough*Amazon*cough*. Possible plots of world domination aside, here are the reviews:
Playstation 3 Eye
Yes, the source is Wikipedia. But the Sony website does not provide very much information on the unit, comparatively. The NUI group (mentioned in the last post) strongly recommends this camera for infrared detection, housing many threads full of tutorials on making it work to that purpose. While it is not the best camera for the task, it is more than adequate, and is the most cost-effective (retailing at $40 USD). I plan on utilizing at least two for my project (if I ever actually build the thing).
Sony Bravia 52-inch LCD
This is by no means a final decision on the TV I will use, nor is it some crazy attempt for me to rake in advertising money (though I have looked at Google Adsense on one occasion...). I simply had to make a decision on the TV, and I typed in 52-inch LCD, and picked the cheapest. Thus, the TV. For the TV, I will have to remove any IR blocking filter, which, if I recall, would only be the white opaque filter. A filter that is not always placed in an LCD anyway. I wouldn't know until I opened up the thing.
DigiKey
Again, not an endorsement, but if MaximumPC felt they were good enough for their LEDs, so do I. According to the NUI group, the LEDs required should be between 40° and 60°, so as to allow the infrared waves to bounce around, but not leak out. An LED that meets the required specifications(of which there are more), is the Osrams SFH-4350, which DigiKey sells.
Tuesday, April 6, 2010
Monday, April 5, 2010
Review of Sources Part I : Concepts
In this post, I will do a bit of a review of some of the sources I will be using for my Senior Innovation project. This is the engineering project I referred to in my first post. My project is to create a multi-touch mall directory.
The materials I will be reviewing in this post are those related to the concept of my design.
Multi-Touch Surfaces: A Technical Guide.
This is one of the most important sources for my project. The document is a collaboration by developers who attended the Bootcamp on Construction & Implementation of Optical Multi-touch Surfaces at Tabletop 2008 in Amsterdam. This document lists the known technologies utilized in multi-touch surfaces at that time. The document also delves into the specifics of each technology, ranging from projection-screen based surfaces, to LCD screens incorporating infrared LEDs. Upon reading the document, I realized that the latter method would be the most cost-effective, low-maintenance, and higher quality option. Cost-effective because an LCD screen of the size required by a mall-directory would run less money than a projector screen of adequate resolution to provide a clear directory. Low-maintenance because projectors would require frequent bulb replacement, cleaning to avoid dust (thus, over-heating), etc. LCD screens could simply be replaced as the need arose. Higher quality because projectors are not typically of a high resolution (at least not high enough for how large mall-directories are), whereas LCD screens are almost all of the required resolution and clarity.
The document also delves into the different LED designs possible for the LCD screen. One method is the use of infrared transceivers, which contain both an infrared emitter, and a sensor to pick up the emissions. This allows for a more accurate form of tracking. Unfortunately, this method requires the LEDs to be laid in an array on the entire back of the screen. As such, this method becomes increasingly more expensive as the size of the screen is increased, becoming nonviable for screens over 32". The other method is to place infrared LEDs all around the bezel of the LCD screen, and any tracking would be done by an infrared sensitive camera.
The document also analyzes the latency of various projectors and filters. As I am not utilizing a projector, latency is not as much of a concern in my design.
Next up: TacTile
TacTile is an LCD-based multi-touch display developed by the Electronic Visualization Laboratory at the University of Illinois at Chicago. The website provides links to a gallery of pictures showing off the display, as well as the documentation of the display. In addition, the source code, API, and some applications are also provided. The programming aspect should be of the greatest use for me, as the multi-touch tracking I would need is right there. I will have to find someone skilled enough at programming to add what is necessary for what my mall-directory concept will entail, of course.
MaximumPC
Now, when you read through this, you will realize that it is using a projector. So why the heck do I have this as a source? Well, it is partly because it goes into the fine details of actually BUILDING the bloody thing. There is still commonality in the process, such as the LED array, the computer, etc. In addition, the article provides links to sources from which materials such as LEDs can be purchased, which is awesome, because I am not so good at finding the LEDs.
NUI Group
MOTHERLOAD. I honestly have not found a better source for DIY touch-surface creation. In fact, if it weren't for my school's requirement of multiple sources, I would review ONLY this site. There is a link to code for touch-surface displays. There is a wiki that details information on applications, libraries, and tutorials for everything involved in the creation of a touch-surface, such as wiring the LEDs. There is also a community forum with threads galore regarding touch-surface technology. Browsing through the Camera-specific heading, I noticed that nearly all of the threads recommend the Playstation 3 Eye as the best camera for touch-surfaces. These suckers are cheap, and GOOD. All they require is a bit of work to allow for infrared detection. As I build the prototype (and hopefully the finished product) of my design, I plan to utilize this group to a great extent so my design may be perfect.
LG 52-inch Mutli-Touch LCD
In my research, I have discovered that it is very difficult to create an LCD touch-surface with a screen larger than 32", due to how fragile it is. While this video does not provide anything directly useful to my design, it does provide a confidence boost that it CAN be done. Encouragement is still an important part of the engineering process.
Quick edit (forgot this one): Janus Displays
This is a demo for a touch display at The Hilton in Chicago. It also ties into Chicago-OHare airport. Sure, it's not a mall-directory, but it IS a directory. I will be looking to this for inspiration in interface design, as the conventional mall-directory design is not ideal for multiple users. Then again, neither is this, but some elements of the design (such as the isometric multi-level map), should prove useful.
The materials I will be reviewing in this post are those related to the concept of my design.
Multi-Touch Surfaces: A Technical Guide.
This is one of the most important sources for my project. The document is a collaboration by developers who attended the Bootcamp on Construction & Implementation of Optical Multi-touch Surfaces at Tabletop 2008 in Amsterdam. This document lists the known technologies utilized in multi-touch surfaces at that time. The document also delves into the specifics of each technology, ranging from projection-screen based surfaces, to LCD screens incorporating infrared LEDs. Upon reading the document, I realized that the latter method would be the most cost-effective, low-maintenance, and higher quality option. Cost-effective because an LCD screen of the size required by a mall-directory would run less money than a projector screen of adequate resolution to provide a clear directory. Low-maintenance because projectors would require frequent bulb replacement, cleaning to avoid dust (thus, over-heating), etc. LCD screens could simply be replaced as the need arose. Higher quality because projectors are not typically of a high resolution (at least not high enough for how large mall-directories are), whereas LCD screens are almost all of the required resolution and clarity.
The document also delves into the different LED designs possible for the LCD screen. One method is the use of infrared transceivers, which contain both an infrared emitter, and a sensor to pick up the emissions. This allows for a more accurate form of tracking. Unfortunately, this method requires the LEDs to be laid in an array on the entire back of the screen. As such, this method becomes increasingly more expensive as the size of the screen is increased, becoming nonviable for screens over 32". The other method is to place infrared LEDs all around the bezel of the LCD screen, and any tracking would be done by an infrared sensitive camera.
The document also analyzes the latency of various projectors and filters. As I am not utilizing a projector, latency is not as much of a concern in my design.
Next up: TacTile
TacTile is an LCD-based multi-touch display developed by the Electronic Visualization Laboratory at the University of Illinois at Chicago. The website provides links to a gallery of pictures showing off the display, as well as the documentation of the display. In addition, the source code, API, and some applications are also provided. The programming aspect should be of the greatest use for me, as the multi-touch tracking I would need is right there. I will have to find someone skilled enough at programming to add what is necessary for what my mall-directory concept will entail, of course.
MaximumPC
Now, when you read through this, you will realize that it is using a projector. So why the heck do I have this as a source? Well, it is partly because it goes into the fine details of actually BUILDING the bloody thing. There is still commonality in the process, such as the LED array, the computer, etc. In addition, the article provides links to sources from which materials such as LEDs can be purchased, which is awesome, because I am not so good at finding the LEDs.
NUI Group
MOTHERLOAD. I honestly have not found a better source for DIY touch-surface creation. In fact, if it weren't for my school's requirement of multiple sources, I would review ONLY this site. There is a link to code for touch-surface displays. There is a wiki that details information on applications, libraries, and tutorials for everything involved in the creation of a touch-surface, such as wiring the LEDs. There is also a community forum with threads galore regarding touch-surface technology. Browsing through the Camera-specific heading, I noticed that nearly all of the threads recommend the Playstation 3 Eye as the best camera for touch-surfaces. These suckers are cheap, and GOOD. All they require is a bit of work to allow for infrared detection. As I build the prototype (and hopefully the finished product) of my design, I plan to utilize this group to a great extent so my design may be perfect.
LG 52-inch Mutli-Touch LCD
In my research, I have discovered that it is very difficult to create an LCD touch-surface with a screen larger than 32", due to how fragile it is. While this video does not provide anything directly useful to my design, it does provide a confidence boost that it CAN be done. Encouragement is still an important part of the engineering process.
Quick edit (forgot this one): Janus Displays
This is a demo for a touch display at The Hilton in Chicago. It also ties into Chicago-OHare airport. Sure, it's not a mall-directory, but it IS a directory. I will be looking to this for inspiration in interface design, as the conventional mall-directory design is not ideal for multiple users. Then again, neither is this, but some elements of the design (such as the isometric multi-level map), should prove useful.
Subscribe to:
Posts (Atom)